End of training

README.md

@@ -1,11 +1,13 @@
 ---
-license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: norallm/normistral-7b-warm
+datasets:
+- hugodk-sch/aftonposten_title_prefs
 model-index:
 - name: ap-normistral-7b-align-scan
   results: []
@@ -16,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ap-normistral-7b-align-scan
 
-This model is a fine-tuned version of [norallm/normistral-7b-warm](https://huggingface.co/norallm/normistral-7b-warm) on the None dataset.
+This model is a fine-tuned version of [data/ap-normistral-7b-sft-qlora](https://huggingface.co/data/ap-normistral-7b-sft-qlora) on the hugodk-sch/aftonposten_title_prefs dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6733
+- Loss: 0.6762
 - Rewards/chosen: -0.0561
-- Rewards/rejected: -0.1040
-- Rewards/accuracies: 0.6121
-- Rewards/margins: 0.0480
-- Logps/rejected: -37.0066
-- Logps/chosen: -33.0038
-- Logits/rejected: 97.9574
-- Logits/chosen: 97.9847
+- Rewards/rejected: -0.0991
+- Rewards/accuracies: 0.5889
+- Rewards/margins: 0.0431
+- Logps/rejected: -36.9578
+- Logps/chosen: -33.0039
+- Logits/rejected: 97.9360
+- Logits/chosen: 97.9657
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": 97.9656753540039,
+    "eval_logits/rejected": 97.93604278564453,
+    "eval_logps/chosen": -33.003868103027344,
+    "eval_logps/rejected": -36.95783615112305,
+    "eval_loss": 0.6761856079101562,
+    "eval_rewards/accuracies": 0.5888704061508179,
+    "eval_rewards/chosen": -0.05606912076473236,
+    "eval_rewards/margins": 0.04306147247552872,
+    "eval_rewards/rejected": -0.09913058578968048,
+    "eval_runtime": 103.6924,
+    "eval_samples": 343,
+    "eval_samples_per_second": 3.308,
+    "eval_steps_per_second": 0.415,
     "train_loss": 0.6461187455561254,
     "train_runtime": 2556.0398,
     "train_samples": 3079,

eval_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 1.0,
-    "eval_logits/chosen": 98.4628677368164,
-    "eval_logits/rejected": 98.43635559082031,
-    "eval_logps/chosen": -32.64057159423828,
-    "eval_logps/rejected": -36.409393310546875,
-    "eval_loss": 0.686401903629303,
-    "eval_rewards/accuracies": 0.5685215592384338,
-    "eval_rewards/chosen": -0.07895812392234802,
-    "eval_rewards/margins": 0.09818949550390244,
-    "eval_rewards/rejected": -0.17714762687683105,
-    "eval_runtime": 103.7748,
+    "eval_logits/chosen": 97.9656753540039,
+    "eval_logits/rejected": 97.93604278564453,
+    "eval_logps/chosen": -33.003868103027344,
+    "eval_logps/rejected": -36.95783615112305,
+    "eval_loss": 0.6761856079101562,
+    "eval_rewards/accuracies": 0.5888704061508179,
+    "eval_rewards/chosen": -0.05606912076473236,
+    "eval_rewards/margins": 0.04306147247552872,
+    "eval_rewards/rejected": -0.09913058578968048,
+    "eval_runtime": 103.6924,
     "eval_samples": 343,
-    "eval_samples_per_second": 3.305,
-    "eval_steps_per_second": 0.414
+    "eval_samples_per_second": 3.308,
+    "eval_steps_per_second": 0.415
 }